Pentagon Now Has Its Own Versions of ChatGPT and Grok

The United States Department of Defense (Pentagon) has officially launched specialised versions of OpenAI’s ChatGPT and xAI’s Grok for use by its 3 million civilian and military personnel. This move is part of the Pentagon’s efforts to integrate generative artificial intelligence (AI) tailored to the needs of its service members.

Both tools, named ChatGPT Mil and Grok for Government, are now part of GenAI.mil. This centralised and secure portal was first launched last year and previously offered access to Google Gemini.

GenAI.mil is specifically designed so that Department of Defense (DoD) employees can access leading commercial AI models without the risk of transmitting sensitive government data through ordinary consumer channels. This official military version receives exemptions from the data collection that is typically difficult to avoid in general consumer technology products.

According to Department of Defense data, this generative AI portal has proven to be highly popular. To date, GenAI.mil has recorded over 1.7 million unique users out of the total 3 million personnel within the department.

Although both are generative AI tools, the Pentagon has positioned ChatGPT Mil and Grok for slightly different functions.

The launch also highlights the absence of Anthropic’s Claude model from the Pentagon’s ecosystem. This follows a dispute after Anthropic refused to provide unlimited use of its AI tool to the Pentagon and insisted on certain safety guardrails. Consequently, the Trump administration previously labelled Anthropic as a supply chain risk, a designation that is currently being challenged in court.

In addition to OpenAI and xAI, the Pentagon has entered into agreements with several other major technology companies to enhance its AI capabilities, including Amazon Web Services (AWS), Microsoft, Nvidia, and Reflection AI.